i hope these new cores are better
Printable View
i hope these new cores are better
You guys are stating the obvious here. Besides, what makes this any different than any other 8 pipe card that can score ~30K in '01 with proper cpu power?Quote:
Originally Posted by saaya
EDIT: I was just trying to show that these cards are capable of decent '01 scores similar to other mid range cards. (although, they will probably need more MHZ than say a 6600GT)
For sure. Even without more cpu Mhz, could likely reach 30-31K just with higher gpu clocks.Quote:
Originally Posted by perry_78
Quote:
Originally Posted by sabrewolf732
How about 11K. Doesn't seem too bad for a single card, considering its only at 648Mhz.
EDIT: Covered up in the pic but cpu speed was set @ 3,010Mhz.
Despite what ATItool says, the default speeds are 680.5 / 684.
1600s are 4 pipes ;)Quote:
Originally Posted by DrJay
other cards benefit from more cpu power, 6600s for example, 1600s dont, not at all...
iirc i got a boost of 50 points in 2k3 going from 2ghz to 2.5ghz a64 power...
and no difference at all in 2k5
Hi ,is it possible to change the voltage through Ati tool for x1600,or only with Ati overclock tool?
...other cards that are in the same mid-range category and happen to be 8 pipes...don't be so literal.Quote:
Originally Posted by saaya
1600s do benefit from extra cpu power...that is if you are talking about '01.
ATItool hasn't worked for me on X1300 or 1600.Quote:
Originally Posted by TheVaLVe
well it doesnt matter if its 4 pipes or 8 pipes, all other mid range card i know scale very well with more cpu power, the 1600s dont.Quote:
Originally Posted by DrJay
the fact that they are 4 pipes probably explains it, its a fast 4 pipe card, but its a 4 pipe card, and other 4 pipe cards are maxed out cpu wise as well by todays cpus.
and of course it beneftis from cpu power in 2k1 :P
2k1 is cpu limited for almost all cards nowadays, hence they will all scale better with higher cpu clocks.
this doesnt mean that the cpu scales well with more cpu power.
its in 2k3 and 2k5 where you can see if and how well cards scale with more cpu power
But 2K1 is what we were talking about. The '01 score I posted yesterday prompted the last several posts.
I have seen about as much increase (from cpu scaling)in scores in '03 and '05 with the 1600xt as with other cards. (not that much)
No problems though, we can disagree. ;)
what cards for example?
i see much bigger bumps with more cpu speed with my other cards...
anyways, i think i found the bios lock!!! :banana:
i tried the latest rabit but it still doesnt work, it doesnt crash when loading the bios but it cant set anything or even read the memory timings.
so i went for a hex editor and looked at the bios i made.
i knew that the reference clock for the card was 27mhz, thats all the extra info rabit gave me.
so i searched for 27
5b2c and following it found this:
27;9743234551,'/0234433258:=<98899:741
98899:741 is what? 133
and whats the speed the 1600 cards stop clocking higher?
630mhz, or to put it differently 500mhz default clock +133mhz = 633mhz
27mhz reference clock ??? =< (is below) 133
could it be that this is basically the lock telling the cards default clock+133mhz = :nono: ?
whoooohooo im getting an unlocked 1600pro and 1600xt bios :banana:
ill post them here as soon as i get them :)
Im not saying the x1600xt is a slow card, its actually a damn decent card, but 03 and 05 are too gpu intensive for the cpu to make much differance imoQuote:
Originally Posted by DrJay
tbh my x800pro didnt really scale at all with cpu power, neither did my 6800 :confused:Quote:
Originally Posted by saaya
Quote:
Originally Posted by saaya
:woot: :toast: let us now how these x1600s can clock! would be crazy if they can hit 750+ :cool:
my x1600xt arrived today.. notice a lot of missing SMD components.
Anyone got links to pics of a Sapphire X1600XT vs. other manu's X1600Pro/XT's sporting GDDR3 so we could compare PCB components (hopefully similar design).
Yeah, that was my point. It is a decent card. Compares nicely to X700s, 6600gt and some others in '01. Scales just as well in '01. Seems to have the same lack of cpu scaling as other cards in '03 and '05.Quote:
Originally Posted by sabrewolf732
I'll be volt modding my card as soon as I get home from work tomorrow. 5K Ohm pot between pins 5 and 7. Stock Vgpu is 1.37.........resistance between FB and GND is 59 Ohms. pot will need to be turned below 2K before any significant gains are realized.
yes, its a nice card, it beats the 6600gt, but it costs more/the same and the 6600s are 2 years old now... so... wow! they beat 6600s... :P
they are ok cards, but im still surprised how huge the core is... the 6600 is tiny and performs around the same...
That is kinda strange. The core on my 1600 is smaller than the NV43 die.
Voltmodded my x1600xt. Stock Vgpu=1.37 / modded Vgpu set @ 1.42.
With room temp water cooling @ 702Mhz core / 837Mhz mem, got the following scores:
01: 30,458
03: 11,561
05: 6,131
Still can't publish X1600 class projects.
Interesting...might have to pick up a second card.
Next stop 750Mhz.
yes but its 90nm vs 110nm, rv530 is bigger than nv43Quote:
Originally Posted by DrJay
http://www.cdrinfo.com/Sections/Arti.../card/nv43.jpg
http://www.itxian.com/Files/BeyondPi...1600_thumb.jpg
cool, so your bios is not locked. what driver did you use?
what app did you use to oc?
what card is it? a sapphire? powercolor? asus?
actually the 6600 and 6800 are alot slower if you look at benchiesQuote:
Originally Posted by saaya
Suspect dr jay got an unlocked bios on his card.
cat 5.13s, Overclocker tool, SapphireQuote:
Originally Posted by saaya
Re: Saaya, come on now....you can look right a NV43 core and plainly see that it is larger than the RV530. You're saying that doesn't count because it is built on a 90nm process? Or, are you just talking about transistor count and not actual size?
As far as the '03 score in my last post goes, that's not really any higher than a 6600GT, is it? Hmmmmm
hmmm did you use the 5.13ccc?
and yeah i was talking about transistor size not size of the core, sorry, should have made myself more clear :)
its mostly about performence/transistors, and rv530 really sux there... :/
especially if you consider its 2 years!!! older than rv530